A scanned or exported PDF is not automatically accessible. Use this practical PDF accessibility checklist to improve OCR, document structure, reading order, headings, alternative text, forms, contrast, and keyboard navigation before sharing a file publicly or with a client.
Overview
PDF accessibility means making a document usable by people with different ways of reading, navigating, and interacting with information. A person may use a screen reader, keyboard navigation, magnification, high-contrast settings, or other assistive technology. Accessibility also improves ordinary usability: clear headings, searchable text, descriptive links, and predictable navigation help almost everyone.
The starting point depends on how the PDF was created. A digitally exported document may already contain selectable text and some structural information. A scan is usually an image inside a PDF and needs OCR before a reader can search or interpret the words. Neither process guarantees a usable reading order or correctly labeled structure.
Use this workflow for reports, instructions, application materials, client deliverables, public records, and other files that people need to read or complete. It is practical guidance, not legal advice. Accessibility obligations can vary by jurisdiction, organization, audience, and document purpose. Where formal conformance is required, compare the finished file with the applicable WCAG guidance, PDF/UA requirements, and your organization’s review process.
Checklist by scenario
For a scanned PDF
- Prepare the source. Scan pages upright, at a readable resolution, with even lighting and minimal background noise. Remove blank pages and check that no page is cropped.
- Run OCR. Use an OCR document scanner or PDF editor to create a text layer. Review the result rather than assuming it is accurate, especially for names, figures, symbols, tables, and handwritten content.
- Check the language. Set the document’s primary language and confirm that unusual words, abbreviations, and quoted passages are handled sensibly by the reading technology available to your audience.
- Set the reading order. Make sure content is read from the intended top-to-bottom and left-to-right sequence. Multi-column pages, sidebars, captions, footnotes, and tables often need manual adjustment.
- Add structure. Tag headings, paragraphs, lists, tables, and decorative elements where the PDF tool supports tagging. A scan may need more manual remediation than a document exported from a structured source.
For a PDF exported from a word processor or design tool
- Use real text. Do not export text as an image unless there is a specific reason. Confirm that readers can select and copy meaningful text.
- Build a heading hierarchy. Use one clear title and a logical sequence of section headings. Do not create headings only by enlarging or bolding ordinary body text.
- Preserve tags and bookmarks. Export settings may include document tags, bookmarks, links, and an initial view. Confirm that these features survive the export.
- Describe meaningful images. Add concise alternative text to charts, diagrams, photos, and icons that communicate information. Mark purely decorative elements as artifacts when appropriate.
- Review tables and lists. Check that table headers are identified and cells are read in a sensible order. Confirm that bullets, numbering, and nested lists remain distinguishable.
For a fillable PDF form
- Label every field. A field should have a clear name or tooltip that explains what the user must enter, such as “Billing address” rather than “Field 7.”
- Set a logical tab order. Keyboard users should move through fields in the same practical sequence as someone reading the form.
- Explain required information. Identify required fields, accepted formats, and any instructions before the user reaches the field. Error messages should describe how to fix a problem.
- Test controls. Check checkboxes, radio buttons, dropdowns, date fields, signature fields, and submit actions with a keyboard and the intended PDF viewer.
If the file will be signed, accessibility should be checked before sending it for signature. A clear signing workflow reduces confusion for every signer. For related process guidance, see the secure contract signing workflow checklist and the guide to adding a signature to a PDF.
What to double-check
Use the following review table when a PDF fails an accessibility check or a reader reports a problem.
| Observed problem | Likely cause | Practical fix |
|---|---|---|
| The screen reader announces no useful text | The file is an image-only scan or OCR failed | Run OCR, correct recognition errors, and confirm that text can be selected and searched |
| Content is read in the wrong order | Reading-order regions or tags are incorrect | Reorder tags or content regions, then test a representative page with multiple columns or sidebars |
| Headings are missing or inconsistent | Visual formatting was used without structural tags | Apply heading levels that reflect the document outline and remove false heading tags |
| An image is announced without useful context | Alternative text is missing, vague, or duplicated | Describe the image’s purpose; if it is decorative, treat it as decorative instead |
| A table is difficult to understand | Header cells, row relationships, or reading order are unclear | Identify headers, simplify complex layouts where possible, and verify how the table is announced |
| A form cannot be completed by keyboard | Fields lack labels or the tab order is arbitrary | Name fields, set a logical tab sequence, and test every control without a mouse |
| Text or controls are hard to distinguish | Color contrast is weak or color is carrying information alone | Adjust colors, add text or symbols, and check contrast using an appropriate testing tool |
Run an automated PDF accessibility checker, but treat its results as a starting point. Automated checks can identify missing tags, metadata, or other detectable patterns; they cannot reliably judge whether the reading order makes sense, whether alternative text conveys the right meaning, or whether instructions are understandable. Perform a manual review as well.
At minimum, test with keyboard navigation, text selection and search, a screen reader where available, zoom or magnification, and more than one PDF viewer if the file is important. Keep the source document and an editable remediation copy so corrections can be made without rebuilding the file from scratch.
Common mistakes
- Stopping after OCR. OCR creates searchable text, but it does not automatically create correct tags, headings, alternative text, or reading order.
- Using color as the only signal. A red label, colored border, or shaded cell may not communicate its meaning to every reader. Add text, symbols, or an explicit instruction.
- Making every visual element a heading. Decorative labels and oversized text can create a confusing navigation outline when tagged incorrectly.
- Leaving the default tab order. A form may look orderly while its fields are reached in an unexpected sequence.
- Writing unhelpful alternative text. “Image” or a filename does not explain why an image is present. Describe its relevant message, or mark it decorative when it adds no information.
- Testing only the original file. Compression, conversion, password protection, signing, or a later edit can change tags, links, fields, or text. Test the exact version you will distribute.
- Ignoring document metadata. Add a meaningful title and confirm the language. Metadata does not replace structural remediation, but it gives users and tools useful context.
Accessibility is also part of file governance. Record the reviewed filename, version, source, remediation date, and reviewer. For documents that move through approvals, a PDF version-control workflow helps prevent an inaccessible draft from being published as the final file.
When to revisit
Use this checklist before every public upload, client delivery, form launch, report release, or major revision. Recheck the file after changing the source layout, replacing images, adding pages, converting formats, applying security settings, or sending it through a signing workflow.
Review your accessibility process when your PDF editor, scanner, OCR engine, document-management system, or publishing platform changes. Tool updates can alter export behavior, tagging, form fields, or reading order. It is also sensible to revisit the process during scheduled compliance reviews, before seasonal planning cycles, and whenever a reader reports difficulty.
Keep a short pre-publication record:
- Source and final PDF filenames, including the version number
- OCR completed and representative pages checked
- Language, title, headings, tags, links, images, tables, and forms reviewed
- Keyboard and manual reading-order checks completed
- Automated checker results reviewed and exceptions documented
- Final file tested after any signing, compression, or password protection step
For the next document, start with the most accessible source available, export carefully, remediate the PDF, and test the final copy. That repeatable workflow is usually more reliable than trying to repair a large collection of image-only PDFs after publication.
